grounded theory: a qualitative research methodology for physical therapy

12
Grounded theory: a qualitative research methodology for physical therapy Linda Resnik Mellion, MS, PT, OCS, and Melissa Moran Tovin, PT, PhD The interest in grounded theory methodology for physical therapy research has grown in recent years. Grounded theory is a qualitative research methodology, initially developed by sociologists Barney Glaser and Anselm L. Strauss. Signi® cant variations in the approach were later introduced by Strauss and Corbin. The intent of grounded theory methodology is to generate a theory relating to a particular situation. Other distinguishing features of this method include the use of open-ended questioning, the synthesis of data collection with analysis, the use of theoretical sampling, and the systematic procedures for data collection, analysis, and theory development. The purposes of this article are to introduce the tenets of the grounded theory methodology, to present examples of how the methodology can be applied to physical therapy research, and to suggest criteria for evaluating grounded theory studies. INTRODUCTION Qualitative research methodologies have been discussed in the physical therapy literature for almost two decades (Parry, 1991; Scully and Shepard, 1983), yet published qualitative research is still uncommon within the ® eld. Although physical therapy educational pro- grams include courses in research methodo- logy, many curricula do not incorporate qualitative methods. As a result, many thera- pists lack an appreciation of the range of qua- litative research methodologies, or the distinct aspects of the various methodologies. This makes it dif® cult for therapists to distinguish between approaches, apply qualitative methods or critically appraise qualitative research. Grounded theory, one of many methodo- logies within the qualitative paradigm, has its own unique language and rules (Charmaz, 2000; Strauss and Corbin, 1998). It is con- sidered by some to be the most comprehensive qualitative research methodology available (Annells, 1997a; Haig, 1995). Physical therapy research utilizing grounded theory was repor- ted as early as 1983 (Scully and Shepard, 1983), but in the following decade only an occasional piece of grounded theory research on physical The grounded theory study that is cited in the examples throughout this paper was conducted by the ®rst author (Linda Resnik Mellion), in partial ful®llment of doctoral degree requirements for Nova Southeastern University, Ft. Lauderdale, FL. The second author (Melissa Tovin), served as metho- dological consultant and doctoral committee member. The authors wish to thank Dennis Hart, PT, PhD, Leah Nof, PT, PhD and Gail Jensen, PT, PhD for consultation and guidance throughout the research study described in this paper. Linda Resnik Mellion , PT, PhD, OCS, Postdoctoral Fellow, Center for Gerontology and Health Services Research, Brown University, Providence, RI 02906. E-mail: [email protected] Melissa Moran Tovin , PT, PhD, Adjunct Assistant Professor, Physical Therapy Progam, Nova Southeastern University, Ft. Lauderdale, FL 33328. E-mail: [email protected] Accepted for publication & Physiotherapy Theory and Practice (2002) 18, 109¡120 # 2002 Taylor & Francis DOI: 10.1080/09593980290058490 Physiother Theory Pract Downloaded from informahealthcare.com by Lunds Universitet on 10/29/14 For personal use only.

Upload: melissa-moran

Post on 28-Feb-2017

218 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Grounded theory: a qualitative research methodology for physical therapy

Grounded theory: a qualitative researchmethodology for physical therapy

Linda Resnik Mellion, MS, PT, OCS, and Melissa Moran Tovin, PT, PhD

The interest in grounded theory methodology for physical therapy research hasgrown in recent years. Grounded theory is a qualitative research methodology,initially developed by sociologists Barney Glaser and Anselm L. Strauss. Signi® cantvariations in the approach were later introduced by Strauss and Corbin. The intentof grounded theory methodology is to generate a theory relating to a particularsituation. Other distinguishing features of this method include the use ofopen-ended questioning, the synthesis of data collection with analysis, the use oftheoretical sampling, and the systematic procedures for data collection, analysis,and theory development. The purposes of this article are to introduce the tenets ofthe grounded theory methodology, to present examples of how the methodologycan be applied to physical therapy research, and to suggest criteria for evaluatinggrounded theory studies.

INTRODUCTION

Qualitative research methodologies have beendiscussed in the physical therapy literature foralmost two decades (Parry, 1991; Scully andShepard, 1983), yet published qualitativeresearch is still uncommon within the ® eld.Although physical therapy educational pro-grams include courses in research methodo-logy, many curricula do not incorporatequalitative methods. As a result, many thera-pists lack an appreciation of the range of qua-litative research methodologies, or the distinctaspects of the various methodologies. This

makes it dif® cult for therapists to distinguishbetween approaches, apply qualitative methodsor critically appraise qualitative research.

Grounded theory, one of many methodo-logies within the qualitative paradigm, has itsown unique language and rules (Charmaz,2000; Strauss and Corbin, 1998). It is con-sidered by some to be the most comprehensivequalitative research methodology available(Annells, 1997a; Haig, 1995). Physical therapyresearch utilizing grounded theory was repor-ted as early as 1983 (Scully and Shepard, 1983),but in the following decade only an occasionalpiece of grounded theory research on physical

The grounded theory study that is cited in the examples throughout this paper was conducted by the®rst author (Linda Resnik Mellion), in partial ful®llment of doctoral degree requirements for NovaSoutheastern University, Ft. Lauderdale, FL. The second author (Melissa Tovin), served as metho-dological consultant and doctoral committee member. The authors wish to thank Dennis Hart, PT,PhD, Leah Nof, PT, PhD and Gail Jensen, PT, PhD for consultation and guidance throughout theresearch study described in this paper.Linda Resnik Mellion, PT, PhD, OCS, Postdoctoral Fellow, Center for Gerontology and Health ServicesResearch, Brown University, Providence, RI 02906. E-mail: [email protected] Moran Tovin, PT, PhD, Adjunct Assistant Professor, Physical Therapy Progam, NovaSoutheastern University, Ft. Lauderdale, FL 33328. E-mail: [email protected] for publication &

Physiotherapy Theory and Practice (2002) 18, 109¡120#2002 Taylor & FrancisDOI: 10.1080/09593980290058490

Phys

ioth

er T

heor

y Pr

act D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y L

unds

Uni

vers

itet o

n 10

/29/

14Fo

r pe

rson

al u

se o

nly.

Page 2: Grounded theory: a qualitative research methodology for physical therapy

therapy related topics was published (Jensen,1990; Jensen, 1992; Yoshida, 1994). In the lastfew years, however, an increased number ofstudies utilizing grounded theory methodologyhave been described in the physical therapyliterature (Backstrom and Dahlgren, 2000;Jensen, Gwyer, Shepard and Hack, 2000; John-son, 2000; Ohman and Hagg, 1998; Shepard,Hack, Gwyer and Jensen, 1999; Stephensonand Wiles, 2000; Stiller, 2000).

In order to be critical consumers of thegrowing amount of grounded theory literature,it is imperative that physical therapists developan appreciation for this approach. The pur-poses of this article are to introduce the tenetsof the grounded theory methodology, topresent examples of how the methodology canbe applied to physical therapy research, and topresent criteria for critical review of groundedtheory research.

The examples provided throughout thisarticle are taken from a study currentlyunderway (Resnik Mellion, 2002). Althoughthis study employed both quantitative andqualitative methodologies, this article will focusonly on the qualitative phase. We will notattempt to describe the project in full, butrather to draw from this research experience toillustrate how grounded theory was applied.

RESEARCH PARADIGMS

Quantitative and qualitative research paradigmshave distinctly different sets of assumptions,terminologies, and methods. The qualitativeapproach is a process of inquiry into social orhuman phenomena, based on building a com-plex, holistic picture (Creswell, 1994). Qualita-tive research takes an integrative, naturalisticapproach to the world, and is thought to betterinform researchers about the complexity ofhuman behavior and social interaction. In con-trast, a quantitative approach is `̀ based upontesting a theory, composed of variables, mea-sured with numbers, and analyzed with statis-tical procedures, in order to determine whetherthe predictive generalizations of the theory holdtrue’’ (Creswell, 1994).

GROUNDED THEORY

Although qualitative research is sometimescategorized under the general term, `̀ descrip-tive research’’ (Portney and Watkins, 1993), itspurpose often extends beyond description toinclude interpretation, prediction and expla-nation. The intent of grounded theory meth-odology is the generation of a theory relatingto a particular situation (Strauss and Corbin,1994). Theory is de® ned as a `̀ set of well-developed concepts related through statementsof relationship, which together constitute anintegrated framework that can be used toexplain or predict phenomena’ ’ (Strauss andCorbin, 1998).

The originators of grounded theory,sociologists Glaser and Strauss, believed thattheories should be grounded in data from the® eld, not based on an a priori theoreticalorientation (Glaser and Strauss, 1967). Agrounded theory identi® es the major con-structs, or categories of a phenomenon, theirrelationships, and the context and process(Becker, 1993). This emphasis on theorydevelopment is the methodology’s most crucialdistinguishing feature. Other distinguishingfeatures of grounded theory include the fusionof data collection and analysis, the shaping ofdata collection by initial data analysis, and thedetailed and systematic procedures for datacollection, analysis, and theorizing.

Grounded theorists may utilize both qua-litative and=or quantitative sources of data indevelopment of theory (Glaser and Strauss,1967). Glaser and Strauss emphasized that theprocess of generating theory was `̀ independentof the kind of data used,’ ’ and saw no con¯ ictin the use of both types of data sources.

RESEARCH QUESTIONS

Grounded theory methodology is an appro-priate choice for addressing research questionsabout complex relationships, clinical situations,or new areas of inquiry. Underlying thegrounded theory approach are three possibleassumptions: 1) all of the concepts pertaining

110 L. R. MELLION AND M. M. TOVIN

Phys

ioth

er T

heor

y Pr

act D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y L

unds

Uni

vers

itet o

n 10

/29/

14Fo

r pe

rson

al u

se o

nly.

Page 3: Grounded theory: a qualitative research methodology for physical therapy

to a given phenomenon have not yet beenidenti® ed, or 2) if concepts have been identi-® ed, then the relationships between the con-cepts are poorly understood or conceptuallyundeveloped, or 3) nobody has asked thisparticular question, so it is as yet impossible todetermine which variables pertain to this areaand which do not (Strauss and Corbin, 1998).

Researchers in physical therapy have usedgrounded theory methodology to develop the-ories explaining physical therapist attitudestoward professional roles (Ohman and Hagg,1998), expert practitioners in physical therapy(Jensen, Gwyer, Shepard and Hack, 2000;Shepard, Hack, Gwyer, and Jensen, 1999), andthe development of professional ethos in phy-sical therapy (Stiller, 2000). This methodologyhas been employed to study patients’ viewsof the home care experience (Stephensonand Wiles, 2000), the experience of hospicepatients (Stephenson and Wiles, 2000), treat-ment techniques for patients with cerebralpalsy (Backstrom and Dahlgren, 2000), andmothers’ perceptions of parenting childrenwith disabilities (Johnson, 2000). The purposeof author’s research is presented in the caseexample that follows.

Case example: groundedtheory research exampleIn our research grounded theory was used todevelop a theory explaining clinician andorganizational characteristics associated withoutcomes effectiveness in the treatment of lowback pain (Resnik Mellion, 2002).

In keeping with the qualitative paradigm,no speci® c hypotheses are formulated at theoutset of a grounded theory study. Instead, atheoretical framework that answers theresearch questions is developed and tested inconjunction with data collection, analysis, andinterpretation (Strauss and Corbin, 1998).Questions must be ¯ exible, open-ended andbroad enough to enable a thorough investiga-tion to be carried out on facets of a phenom-enon, while providing adequate focus for theresearcher (Charmaz, 1990; Smith and Biley,1997; Strauss and Corbin, 1998).

Case example: researchquestionsThe questions that guided grounded theorygeneration in our study (Resnik Mellion, 2002)were ¯ exible and open-ended, and included:

° Does clinician knowledge and experienceaffect the outcomes of care, if so how?

° Does the clinician’s philosophy of practiceaffect the outcomes of care, if so how?

° What characteristics distinguish betweenhighly effective and effective clinicians?

LITERATURE REVIEWIN GROUNDED THEORY

In a grounded theory approach, general read-ing of the literature is performed to enable theresearcher to approach the subject with somebackground knowledge and to assist theresearcher in formulating questions for initialinterviews and observations (Cutcliffe, 2000;Strauss and Corbin, 1998). Extensive pre-liminary literature review, however, is delayedin the grounded theory approach. Backgroundliterature review should not be too extensive, asthe theories should evolve from the datathemselves, producing a theory grounded inthe data, thus the term grounded theory. Thishelps to insure that data analysis is based in thedata and that pre-existing constructs do notshape the subsequent theory formation. Whenexisting theoretical constructs are used as aframework for research, they must be usedjudiciously so that they can inform, but notdictate, data analysis. (Morse, 2002)

Reading and integrating literature aredelayed in this approach, not omitted, and areregarded as an important part of theorydevelopment (Charmaz, 1990). Delaying theliterature review is thought to decrease thelikelihood that the researcher will be `̀ lockedinto preconceived conceptual blinders’’ and sowill be able to approach the ® eld more openly(Charmaz, 1990). Literature is reviewed con-tinuously throughout data collection andanalysis. The researcher then checks the keyconcepts that have been uncovered against the

GROUNDED THEORY 111

Phys

ioth

er T

heor

y Pr

act D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y L

unds

Uni

vers

itet o

n 10

/29/

14Fo

r pe

rson

al u

se o

nly.

Page 4: Grounded theory: a qualitative research methodology for physical therapy

literature. Concepts derived from the literaturecan provide a source for making comparisonsof the data (Strauss and Corbin, 1998). It isthen possible for the researcher to determinewhether the theory already exists and, if so,what others have said about it (Strauss andCorbin, 1998).

SAMPLING IN GROUNDEDTHEORY

Participants in a grounded theory study arechosen by theoretical sampling, based upontheir expert knowledge of the phenomenonunder scrutiny (Strauss and Corbin, 1998).Theoretical sampling is de® ned as data gath-ering, driven by concepts derived from theevolving theory and based on the conceptsof making comparisons (Strauss and Corbin,1998). In theoretical sampling, the analystsimultaneously collects, codes, and analyzesdata and decides what subsequent data to col-lect in order to develop theory as it emerges.The purpose is to gather data that will max-imize opportunities todiscover variations amongconcepts and to densify categories in terms oftheir properties and dimensions. Theoreticalsampling allows the emerging theory and theanalyst to control the process of data collection.Thus, it requires ¯ exibility in determining theprecise number of subjects, the kinds of inter-view questions or other data that is collectedand the number and types of follow-up.

Some authors of qualitative research usethe term purposeful sampling interchangeablywith theoretical sampling (Cutcliffe, 2000).Others differentiate theoretical from purpose-ful sampling (Cutcliffe, 2000). Purposeful sam-pling involves a calculated decision to sample aspeci® c area because of a preconceived initialset of dimensions, whereas theoretical samplingis driven by the emerging theory (Cutcliffe,2000). Strauss and Corbin make no such dis-tinction in terminology, but do describe twophases of sampling that occur sequentially inthe research process (Strauss and Corbin,1998). According to Strauss and Corbin,the researcher makes some initial sampling

decisions regarding the group to be studied, thetypes of data to be used, the length of study, andthe number of observations and=or interviews.Later, the researcher may modify these deci-sions according to the evolving theory.

In grounded theory the researcher con-tinues sampling until nothing new is being saidabout the concepts under exploration, and thecollected data have reached saturation point(Cutcliffe, 2000; Glaser and Strauss, 1967).Strauss and Corbin point out that saturation isalways a matter of degree, stating there is alwaysthe potential for the `̀ new’’ to emerge (Straussand Corbin, 1998). According to Strauss andCorbin (1998), saturation is considered thepoint in research where collecting additionaldata seems counterproductive in that the newthat is uncovered does not add that much moreexplanation, or the researcher runs out oftime, money or both.

Case example theoreticalsamplingInitial subject selection

Participants were selected following quantita-tive analysis of an outcomes database thatidenti® ed highly effective and effective thera-pists based on patient outcomes. Therapistswere selected from each group because theycould best speak to the phenomenon understudy. Clinicians were sent letters of introduc-tion, and informed consent.

An initial guided interview format wasdeveloped, and a procedure for other forms ofdata collection was established. After theseinitial decisions were made, subsequent datacollection and sampling of those therapists whohad agreed to participate was driven by thedata analysis, and emerging theory. Therapistsfrom each group were contacted in the generalorder in which responses were received.

Theoretical sampling guided by dataanalysis

When data analysis suggested a relationshipbetween work experience prior to physicaltherapy school and therapist group membership

112 L. R. MELLION AND M. M. TOVIN

Phys

ioth

er T

heor

y Pr

act D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y L

unds

Uni

vers

itet o

n 10

/29/

14Fo

r pe

rson

al u

se o

nly.

Page 5: Grounded theory: a qualitative research methodology for physical therapy

(highly effective or effective), questions aboutthese experiences were added to interviews,and=or pursued in follow-up.

Another example of theoretical samplingwas the way that qualitative data analysis shapedfurther quantitative analysis of the dataset. Thediscovery of several relatively new graduateswithin the highly effective group led theresearcher to examine the years of experienceof therapists within each of the selected groups(highly effective and effective therapists) andto compare the years of experience betweengroups. The ® ndings were then used as anadditional source of information for theorydevelopment.

GROUNDED THEORY DATAANALYSIS

Grounded theory methodology involves simul-taneous data collection and several phases ofanalysis performed in systematic steps througha variety of techniques and procedures. Theo-retical concepts are developed from the data ateach of these phases through line-by-line cod-ing and the constant comparative analysismethod. The components of data analysis arediscussed below.

Grounded theory data recordingThe ® rst procedure of grounded theorymethodology is data recording. The data col-lection and the initial analysis should runsimultaneously, wherever possible, with thewriting of ® eld notes or tape recordings tran-scribed immediately after the event. This con-trasts with methodologies that allow completionof data collection and recording prior to dataanalysis. This intertwining of data collection,recording, and analysis is necessary to facilitatetheoretical sampling and the successful gen-eration of grounded theory.

Grounded theory microanalysisOnce suf® cient data have been collected andtranscribed, the researcher begins the processof microanalysis. Microanalysis, which includes

open and axial coding, entails detailed line-by-line analysis used to generate initial categoriesand to suggest relationships among categories(Strauss and Corbin, 1998). Two key proce-dures, asking questions and making compar-isons, guide analysis and aid theorizing.

The constant comparative method is cen-tral to the data analysis in generating groundedtheory (Glaser and Strauss, 1967). Using thismethod, all codes are compared repeatedlywithin and between each other until the basicproperties of a category or construct arede® ned. Coding is considered to be a dynamicand ¯ uid process, but is broken down in Straussand Corbin’s version of the methodology intothree levels of activity: open, axial and selectivecoding (Strauss and Corbin, 1998). Otheranalytic devices, including memo writing,matrix building and the use of diagrams, arealso incorporated as parts of the analysis(Strauss and Corbin, 1998).

Open codingInitial indexing, or coding, begins with line-by-line analysis of all the facets that the researcher® nds important or interesting in the text(Strauss and Corbin, 1998). Open codingdescribes the process through which conceptsare identi® ed and their properties anddimensions are discovered in the data. A con-cept is a labeled phenomenon, an abstractrepresentation of an object, action=interaction,or event that is signi® cant in the data.

When a code is identi® ed, it is recordedwith a short description of the noted occur-rence and its position in the text. The processcontinues by checking the rest of the text for allpossible instances of new codes. The researchercompares incident to incident with the purposeof establishing the underlying uniformity andits varying conditions (Strauss and Corbin,1998). Codes must ® t the phenomenondescribed in the data exactly, so they may needto be continually ® ne-tuned until the ® t isaccurate. Grounded theorists may use hand-written memoing and coding systems, and=oremploy a variety of software developed forqualitative researchers.

GROUNDED THEORY 113

Phys

ioth

er T

heor

y Pr

act D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y L

unds

Uni

vers

itet o

n 10

/29/

14Fo

r pe

rson

al u

se o

nly.

Page 6: Grounded theory: a qualitative research methodology for physical therapy

As the coding progresses, the researchertypically derives dozens of concepts. Usingthe constant comparison analysis method, thecoded concepts are re® ned, extended and cross-referenced with the data as a whole and relatedto each other. Once concepts begin to accu-mulate, the analyst begins the process ofgrouping or categorizing them with moreabstract explanatory terms called categories.

Categories depict the problems, issues,concerns, and matters that are important tothose being studied. Once a category is iden-ti® ed the analyst begins to develop it in termsof its properties and dimensions and differen-tiating the subcategories. Subcategories answerquestions about the phenomenon such as when,why, how, and with what consequences (Straussand Corbin, 1998).

Case example: open codingIn our research, the process of open coding ofthe data from therapist case studies resulted inover 100 different codes.

Overarching categories:

After repeated re® nement, the codes weregrouped into these four overarching categories,each containing numerous subcategories:

° Knowledge base° Current practice° Values and virtues° Clinical reasoning.

Subcategories:

Each overarching category had numerous sub-categories. The subcategories of knowledge, forinstance, included:

° Domains of personal knowledge° Craft knowledge° Formal knowledge.

Each subcategory had its own subcategories.Personal knowledge, for example, included:

° Work experience prior to physical therapyschool

° Diversity of experience

° Experience with movement and sports° History as patient° Life experience with pain and disability° Knowledge of people.

Axial codingCoding is used to gain an understanding ofa phenomenon and discern relationshipsbetween categories. Axial coding continues theprocess begun in open coding, of relatingcategories to their subcategories, and linkingcategories at the level of properties anddimensions (Strauss and Corbin, 1998). Openand axial coding are not sequential acts, butproceed together in a ¯ uid process throughoutthe analysis.

Strauss and Corbin (1998) recommend theuse of an organizational scheme they call theparadigm model, to establish relationshipsbetween the data during axial coding. The useof this model is one of the key differentiatingfeatures between Glaser and Strauss’s versionof grounded theory (Kendall, 1999). Thecomponents of the paradigm model includethe identi® cation of phenomena, conditions,actions=interactions, and consequences. Incoding, the categories stand for phenomena.

Conditions are considered sets of events orhappenings that create the situations, issues,and problems pertaining to phenomena. Con-ditions explain to some degree why and howpersons or groups respond in certain ways(Strauss and Corbin, 1998). Labels placed onconditions such as causal, intervening, andcontextual are ways of sorting and distinguish-ing some of the complex relationships amongconditions and their relationships to actionsand interactions.

Actions and interactions denote the activ-ities amongst individuals, groups, and organi-zations and include discussions, negotiations,and actual actions. Strategic actions and inter-actions are purposeful or deliberate acts thatare taken to resolve a problem and which shapethe phenomenon in some way (Strauss andCorbin, 1998). Routine actions and interac-tions are more habituated ways of respondingto everyday occurrences and may include rules,

114 L. R. MELLION AND M. M. TOVIN

Phys

ioth

er T

heor

y Pr

act D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y L

unds

Uni

vers

itet o

n 10

/29/

14Fo

r pe

rson

al u

se o

nly.

Page 7: Grounded theory: a qualitative research methodology for physical therapy

regulations, policies, procedures or establishedprotocols (Strauss and Corbin, 1998). The ® nalcomponent of the paradigm model is con-sequences, which might be intended or unin-tended, immediate or cumulative, reversible orirreversible, foreseen or unforeseen.

The paradigm model is a device to assistinvestigators in systematically developing andrelating concepts and categories to each other,but it need not be applied rigidly (Strauss andCorbin, 1998). Grounded theorists who do notadhere to the Strauss and Corbin methodologywould not employ this model, but would allowthe analysis of relationships to emerge in a lessstructured way through the use of questioningand constant comparison.

Case example: axial codingprocess° Summary statements of case analyses were

compared between therapists, and thenbetween the two groups of therapists. Thisconstant comparison helped to clarify thetrends and patterns of interpretations andrelationships.

° Axial coding of the key categories pro-ceeded with identi® cation of the properties,dimensions and relationships of eachcategory.

° The literature was consulted to compare® ndings and emerging explanations to exist-ing research and theories.

Fig. 1 Cross case comparison highly effective novice and experienced groups. This schematic identi®es the keyproperties and dimensions of the overarching categories and their hypothesized relationships.

GROUNDED THEORY 115

Phys

ioth

er T

heor

y Pr

act D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y L

unds

Uni

vers

itet o

n 10

/29/

14Fo

r pe

rson

al u

se o

nly.

Page 8: Grounded theory: a qualitative research methodology for physical therapy

° The paradigm model was applied duringthis data analysis and conditions, actions andinteractions and consequences were hypo-thesized.

° The ® nal diagram comparing highly effec-tive novice and experienced therapists isshown in Figure 1.

Case example axial coding:conditionsA brief summary of the analysis of the highlyeffective novice phenomenon illustrates howaxial coding unfolded.

° An early ® nding was that number of years ofclinical experience, a dimension of knowl-edge base, did not distinguish the highlyeffective from the effective therapists. There-fore, years of clinical experience was notconsidered to be a condition of high effec-tiveness levels.

° Several other conditional relationships wereobserved. Each highly effective novicetherapist had an undergraduate degree inexercise science, prior work experience asphysical therapy aide or athletic trainer, andpersonal experience as an athlete and aphysical therapy patient.

° An excerpt of an interview with one highlyeffective novice, who spent four years as arehab aide in a geriatric setting, and fouryears as an athletic trainer in an outpatientorthopaedic setting before entering physicaltherapy school illustrates this conditionalrelationship.

Well obviously being with the geriatricpopulation, it helped me to learn andunderstand how they are, how to dealwith them, just they are different thenworking with middle-aged and kids,and you treat them differently than youtreat high school kids vs. middle agedpeople, adults, vs. a geriatric popula-tion. I think I’ve had a wide variety ofpatient contact, that’s pretty muchhelped me to be a well-rounded personin treating all populations age-wise.

Case example axial coding:actions and interactions° There were indications that therapists in the

highly effective group consciously thoughtabout and analyzed their practice more thantherapists in the effective group (actions andinteractions). Here is an example of datafrom a highly effective novice therapist, whodescribed her strategy for integrating newmaterial into clinical practice.

I’m going to pick one patient todaythat I’m going to apply what I’velearned and take the time to do, so atleast feel good that I’ve done that. Anda lot of time someone would pop up,where I go, I learned about this fora minute, I’d take the time to stop,go look this up, apply it and thinkthrough the whole process you know.So then I started to gradually use eachlittle thing that I’d learn, so now then Ifelt I knew more of what I was doing,but it just took time though.

° Therapists in the highly effective groupreported that they frequently conferredwith colleagues and utilized their peers to`̀ take a peek’’ at challenging patients (actionsand interactions). This is illustrated by theremarks made by another highly effectivetherapist.

It’s a really strong staff and for meespecially, the strongest thing I canthink of is that they are all very willingto answer questions. They are willing tocome over and take a peek at a patientthat I’ ve been having a hard time with.

° Highly effective therapists seemed moreinquisitive and more cognizant of their ownlimitations. As one highly effective noviceacknowledged, `̀ I know that I’ve got a lot tolearn.’’ He explained how he sought answersby `̀ constantly asking questions.’ ’ These qual-ities were hypothesized to result in greaterself-re¯ ection and metacognition, andincreased dialogue and re¯ ection with col-leagues (actions and interactions).

116 L. R. MELLION AND M. M. TOVIN

Phys

ioth

er T

heor

y Pr

act D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y L

unds

Uni

vers

itet o

n 10

/29/

14Fo

r pe

rson

al u

se o

nly.

Page 9: Grounded theory: a qualitative research methodology for physical therapy

Case example axial coding:consequencesThe result (consequence) of highly effectivenovice therapists’ values and virtues and beha-vior was theorized to be an accelerated acqui-sition of knowledge.

Selective codingThe ® rst step in theory integration is deter-mining the central category that represents themain theme of the research. The central cate-gory has explanatory, analytic power. Selectivecoding is the term used by Strauss and Corbin(1998) to describe this process of integratingand re® ning theory. They discuss several tech-niques for helping to identify the central cate-gory, including the use of storyline writing,memo analysis, integrative diagrams, and reviewof the literature.

Once the central category is identi® ed, thetheory must be re® ned through reviewing forinternal consistency, gaps in logics, supple-menting any poorly developed categories, andreducing excess categories that do little tocontribute to the theory (Strauss and Corbin,1998). Since the analyst is searching for densityof the properties and dimensions of a category,poorly developed categories can be saturatedthrough additional theoretical sampling. Thelast step in theory building is validation, whichis done by comparing it to the raw data and=orby presenting the theory to participants fortheir reactions (Strauss and Corbin, 1998).Theory integration occurs over the course ofanalysis and may not end until the ® nal writing(Charmaz, 2000; Strauss and Corbin, 1998).

Case example: theorydevelopmentMemo analysis, diagrams, review of the litera-ture, analysis of the case studies, matrices andthe quantitative dataset were used to furtherre® ne and develop a theory. Full presentationof this theory is beyond the scope of this paper,however, several key aspects are included todemonstrate how the discovery of highlyeffective novices was handled.

° `̀ Highly effective’’ clinicians possessed a mul-tidimensional knowledge base that includedentry-level education, continuing educa-tion, personal knowledge, specialty work,clinical experience, and pooled collegialknowledge.

° The acquisition of knowledge is facilitatedby the use of re¯ ection and by an organiza-tional culture that supports re¯ ective prac-tice and professional growth.

° The individual components (e.g., years of cli-nical experience) of this knowledge base arenot as critical as the sum total of knowledge.

° The highly effective practitioner’s knowl-edge base often begins with academicdiversity prior to physical therapy schooland=or an undergraduate degree in exercisescience.

° The clinician’s knowledge acquisition andorganization may also be facilitated by workexperience prior to physical therapy schooland continued clinical experience.

° The clinician’s personal knowledge ofmovement and rehabilitation serves as aknowledge source. This personal knowledgeis gleaned through participation in athletics,personal injury, and the experience of beinga patient.

° Other important sources of knowledge mayinclude knowledge of psychosocial factors inrehabilitation and knowledge of teaching.These sources facilitate effective patient com-munication and education.

° Another important source of knowledge isspecialty education, speci® cally manualtherapy certi® cation, and residency training.

° Knowledge is also gained from the currentpractice environment, opportunities to spe-cialize in the clinic, shared knowledge, andre¯ ection amongst colleagues and oppor-tunities for growth provided in the workenvironment.

° The practitioner’s values and virtues areinstrumental in utilizing and gaining know-ledge. Clinicians who value and appreciatepatient individuality are able to garner moreinformation from and about patientsthrough attentive listening, trust buildingand observation.

GROUNDED THEORY 117

Phys

ioth

er T

heor

y Pr

act D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y L

unds

Uni

vers

itet o

n 10

/29/

14Fo

r pe

rson

al u

se o

nly.

Page 10: Grounded theory: a qualitative research methodology for physical therapy

° Clinician knowledge is also increasedthrough a love of learning, a drive to con-tinually improve, and a commitment toprofessional growth.

° The use of self-monitoring and re¯ ection isinstrumental in learning from practice.

CRITERIA FOR EVALUATINGGROUNDED THEORY RESEARCH

Qualitative researchers utilize the terms cred-ibility, transferability,dependability andcon® rm-ability in place of the terms internal validity,external validity, reliability and objectivity toevaluate the quality of research (Lincoln,1985). Qualitative researchers also rede® ne themeaning of key concepts used in researchevaluation (Strauss and Corbin, 1998). Repro-ducibility, for example, in quantitative researchmeans that a study can be replicated throughthe same or similar research processes.However, this is unlikely to hold true for aqualitative study, which examines a social phe-nomenon in the real world. In a qualitativestudy reproducibility means that `̀ given thesame theoretical perspective of the originalresearcher, following the same general rules fordata gathering and analysis, and assuming asimilar set of conditions, other researchersshould be able to derive either the same or asimilar theoretical explanation for the phe-nomenon under investigation’’ (Strauss andCorbin, 1998).

Numerous strategies are used in the ® eldof qualitative research that enhance the trust-worthiness or rigor of a qualitative study. Theseinclude prolonged engagement with subjects,persistent observation, triangulation of datasources, peer reviews=debrie® ng, negative caseanalysis, clari® cation of researcher bias, mem-ber checks, thick description and the use ofexternal audits (Lincoln, 1985).

In addition, widely accepted standards thatexist for evaluating scienti® c theory (Gortnerand Schultz, 1988), are rede® ned to ® t therealities of qualitative research and the com-plexities of studying social phenomena (Haig,

1995; Strauss and Corbin, 1998). For instance,in qualitative research, the standard of gen-eralizability of ® ndings is rede® ned. Groundedtheorists consider the explanatory power ofa theory rather than it’s generalizability(Strauss and Corbin, 1998). Explanatory poweris the predictive ability to explain what mighthappen in given situations and what con-sequences might occur as a result of actionsand interactions.

In quantitative research, theories are ver-i® ed or validated through testing. This pre-sumes that theories have `̀ arisen full blown’’and are ready for testing (Haig, 1995). Incontrast, grounded theorists hold a dynamicperspective on theory as ever developing andnot as a perfected product (Haig, 1995).

Case example strategiesto enhance trustworthinessOur study employed a variety of strategies toenhance trustworthiness.

° Source triangulation of data from multiplesources (including the outcomes database,written statements of philosophy, partici-pant’s curriculum vitae as well as interview)was used.

° Member checks allowed several researchparticipants to review research materials inorder toverify the researcher’s interpretation.

° Thick descriptions in the form of mean-ingful quotations to represent importantthemes and categories were included whenreporting data analysis.

° Peer review and debrie® ng after case con-struction was performed, providing anopportunity for review and comment onthe data.

° A re¯ exive journal was used to documentthe researcher’s thoughts, feelings, anddecision-making throughout the data col-lection and analysis process.

° The researcher’s bias was examined andclari® ed. Finally, an external audit was con-ducted to review the data and assist in clar-i® cation of the researcher’s perspective.

118 L. R. MELLION AND M. M. TOVIN

Phys

ioth

er T

heor

y Pr

act D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y L

unds

Uni

vers

itet o

n 10

/29/

14Fo

r pe

rson

al u

se o

nly.

Page 11: Grounded theory: a qualitative research methodology for physical therapy

Evaluative questionsWhen evaluating any qualitative research, thereader should determine the strategies used forenhancing rigor (Lincoln, 1985). These shouldbe described in published research. In addi-tion, Strauss and Corbin (1998) suggest the useof a speci® c set of questions for evaluating theadequacy of the grounded theory researchprocess and the empirical grounding of theresearch.

1) Are concepts generated? Do theoreticalbuilding blocks come from the data or someother source?

2) Are the concepts systematically related?3) Are there many conceptual linkages, Are the

categories well developed?4) Do categories have conceptual density?5) Is variation built into the theory?6) Are the conditions under which variation

can be found built into the study andexplained?

7) Has process been taken into account?8) Do the theoretical ® ndings seem signi® cant

and to what extent?9) Does the theory stand the test of time and

become part of the discussions and ideasexchanged among relevant social and pro-fessional groups?

CONCLUSION

Since its introduction in the 1960s, groundedtheory has developed to be one of the mostcomprehensive qualitative research methodol-ogies available. Although grounded theory is aqualitative research methodology, it can utilizeboth quantitative and qualitative data sources.Key features of grounded theory methodologyinclude the use of open-ended research ques-tions, delayed review of the literature, andtheoretical sampling.

This methodology is appropriate for investi-gating many topics of relevance to physicaltherapy. It is particularly appropriate wheninvestigating new areas, where it is not possibleto determine which variables pertain to anarea and which do not, or when relationships

between concepts are poorly understood orconceptually undeveloped.

As physical therapy researchers continue todevelop sophistication and familiarity with thismethodology, it is likely that the amount ofphysical therapy research utilizing a groundedtheory approach will continue to expand. Inthis article we explained how we appliedgrounded theory methodology, drawing fromboth quantitative and qualitative data sources,to develop a theory explaining therapist effec-tiveness in the treatment of patients with lowback pain. The adequacy of the resulting the-ory has yet to be evaluated by our peers.

ReferencesAnnells M 1997a Grounded theory method, part I: within

the ® ve moments of qualitative research. Nursing Inquiry4: 120¡129

Annells M 1997b Grounded theory method, part II: optionsfor users of the method. Nursing Inquiry 4: 176¡180

Backstrom B, Dahlgren L 2000 Vojta self-training: experi-ences of six neurologically impaired people: a qualitativestudy. Physiotherapy 86: 567¡574

Becker PH 1993 Common pitfalls in published groundedtheory research. Qualitative Health Research 3: 254¡260

Charmaz K 1990 `̀ Discovering’’ chronic illness: usinggrounded theory. Social Science & Medicine 30:1161¡1172

Charmaz K 2000 Grounded theory: objectivist and con-structivist methods. In: Denzin N, Lincoln Y (eds)Handbook of qualitative research. pp 509¡537 SagePublications Ltd., Thousand Oaks, CA

Creswell J 1994 Combined qualitative and quantitativedesigns research design: qualitative & quantitativeapproaches. Sage Publications, Thousand Oaks, CA

Cutcliffe JR 2000 Methodological issues in grounded the-ory. Journal of Advanced Nursing 31: 1476¡1484

Glaser B, Strauss A 1965 Awareness of dying. Aldine,Chicago

Glaser B, Strauss A 1967 The discovery of grounded theory:strategies for qualitative research. Aldine, Chicago

Gortner SR, Schultz PR 1988 Approaches to nursing sci-ence methods. Image ¡ the Journal of Nursing Scho-larship 20: 22¡24

Haig B 1995 Grounded Theory as Scienti® c Method.Available: http:==www.ed.uiuc.edu= EPS=PES-Yearbook=95_docs=haig.html Accessed 14=5=01

Jensen GM, Gwyer J, Shepard KF, Hack LM 2000 Expertpractice in physical therapy . . . including commentaryby Bartlett D, Mitchell RU and Swaine BR with authorresponse. Physical Therapy 80: 28¡52

Jensen GM, Shepard KF, Gwyer J, Hack LM 1992 Attributedimensions that distinguish master and novice physicaltherapy clinicians in orthopedic settings. Physical Therapy72(10): 711¡722

Jensen GM, Shepard KF, Hack LM 1990 The novice versusthe experienced clinician: insights into the work of thephysical therapist. Physical Therapy 70(5): 314¡323

GROUNDED THEORY 119

Phys

ioth

er T

heor

y Pr

act D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y L

unds

Uni

vers

itet o

n 10

/29/

14Fo

r pe

rson

al u

se o

nly.

Page 12: Grounded theory: a qualitative research methodology for physical therapy

Johnson BS 2000 Mothers’ perceptions of parenting chil-dren with disabilities. MCN American Journal ofMaternal Child Nursing 25: 127¡132

Kendall J 1999 Axial coding and grounded theory con-troversy. Western Journal of Nursing Research 21:743¡757

Lincoln YS, Guba EG 1985 Naturalistic inquiry. Sage Pub-lications, Beverly Hills, CA

Morse JM 2002 Theory innocent or theory smart? Quali-tative Health Research 12(3): 295¡296

Ohman A, Hagg K 1998 Attitudes of novice physiotherapiststo their professional role: a gender perspective. Phy-siotherapy Theory & Practice 14: 23¡32

The Origin of Grounded Theory Methodology. Available:http:==www.geocities.com=ResearchTriangle=Lab=1491=gtm-i-asal.html Accessed 28=5=2001

Parry A 1991 Physiotherapy and methods of inquiry:con¯ ict and reconciliation. Physiotherapy 77: 435¡438

Portney L, Watkins M 1993 Foundations of clinical researchapplications to practice. Appleton & Langer, Stamford,CN

Resnik Mellion L 2002 Clinician and organizational factorsassociated with outcomes effectivenes s in the treatmentof patients with low back pain. Unpublished doctoraldissertation, Nova Southeastern University

Scully RM, Shepard KF 1983 Clinical teaching in physicaltherapy education. An ethnographic study. PhysicalTherapy 63: 349¡358

Shepard KF, Hack LM, Gwyer J, Jensen GM 1999 Describingexpert practice in physical therapy. Qualitative HealthResearch 9: 746¡758

Smith K, Biley F 1997 Understanding grounded theory:principles and evaluation. Nurse Researcher 4: 17¡30

Stephenson S, Wiles R 2000 Advantages and disadvantagesof the home setting for therapy: views of patients andtherapists. British Journal of Occupational Therapy 63:59¡64

Stiller C 2000 Exploring the ethos of the physical therapyprofession in the United States: social, cultural, and his-torical in¯ uences and their relationship to education.Journal of Physical Therapy Education 14: 7¡15

Strauss A, Corbin J 1994 Grounded theory methodology: anoverview. In: Denzin N, Lincoln Y (eds) Handbook ofqualitative research. Sage, Thousand Oaks, CA

Strauss A, Corbin J 1998 Basics of qualitative research:techniques and procedures for developing gro-unded theory. 2nd ed. Sage Publications, ThousandOaks, CA

Yoshida KK 1994 Institutional impact on self conceptamong persons with spinal cord injury. InternationalJournal of Rehabilitation Research 17: 95¡107

120 L. R. MELLION AND M. M. TOVIN

Phys

ioth

er T

heor

y Pr

act D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y L

unds

Uni

vers

itet o

n 10

/29/

14Fo

r pe

rson

al u

se o

nly.